DESCRIPTION: (Applicant's Description) This is a prospective randomized clinical trial designed to study the biologic effects of combination biochemopreventive therapy in the reversal of laryngeal dysplasia and the effects of fenretinide maintenance therapy in maintaining or obtaining further response. In the upper aerodigestive tract advanced premalignant lesions harboring substantial genetic changes are not effectively reversed with retinoids alone. The combination of 13-cis retinoic acid (13cRA), alpha tocopherol, and alpha-interferon has produced dramatic reversal of the majority of the laryngeal dysplastic lesions treated in a currently ongoing study, but certain genetic changes persisted despite complete response. The hypotheses underlying this proposal are: a) Moderate to severe dysplasia in laryngeal epithelium results from chronic carcinogen exposure and may serve as a model for developing chemoprevention strategies. b) Biomarkers may be used to define the molecular and cellular changes associated with field and multistep carcinogenesis and for cancer risk assessment. c) Biochemopreventive therapy will be effective in reversing moderate to severe laryngeal dysplasia and will alter biomarker expression. d) Fenretinde will be effective in response maintenance and in maintenance of biomarker modulation with less toxicity than induction therapy and will be more effective than placebo. e) Fenretinide will reverse persistent lesions and will further modulate and possibly reverse genotypic and phenotypic alterations not already modulated or reversed by induction therapy. The following objectives are proposed: a) To confirm the efficacy and toxicity of combination biochemoprevention for 12 months in patients with moderate to severe laryngeal dysplasia. b) To compare the efficacy of fenretinide in maintaining or achieving further response with a placebo control and determine its toxicity. c) Evaluate the effects of biochemoprevention on the expression of genetic biomarkers (chromosome polysomy, loss of heterotozygosity at chromosomes 9p, 3p and 17p, p53, p16 and cyclin Dl gene status and protein expression), phenotypic change biomarkers (proliferation: Ki-67), and nuclear retinoic acid receptor expression. d) Evaluate the efficacy of maintenance therapy in maintaining reversal and or reversing persistent genotypic and phenotypic abnormalities. All patients (a target sample size of 100) will receive combination biochemopreventive therapy for 12 months, and those showing response or stable disease will be randomized to fenretinide maintenance versus placebo for 24 months. For the biomarker studies, paraffin-embedded biopsy specimens at baseline, 12 months, and 36 months will he used. We will analyze genetic instability by chromosome in situ hybridization; LOH of 9p, 3p, and 17p by microsatellite analysis; p53 alterations by immunohistochemistry and direct sequencing analysis; and alterations in ki-67, p16, and cyclin Dl expression by immunohistochemistry. The study of biomarkers will enhance our understanding of carcinogenesis and of mechanisms of action of the intervention; and, it will improve our ability for risk assessment. The proposed research will eventually direct future intervention trials in premalignancy, not only in the laryngeal setting but in epithelial carcinogenesis in general.